CN111125527B

CN111125527B - Group instance object acquisition method and device based on user matching degree

Info

Publication number: CN111125527B
Application number: CN201911347662.5A
Authority: CN
Inventors: 殷晓明
Original assignee: Koubei Shanghai Information Technology Co Ltd
Current assignee: Koubei Shanghai Information Technology Co Ltd
Priority date: 2019-12-24
Filing date: 2019-12-24
Publication date: 2024-02-27
Anticipated expiration: 2039-12-24
Also published as: CN111125527A

Abstract

The invention discloses a method and a device for acquiring a group instance object based on user matching degree. The method comprises the following steps: extracting user characteristics of a target user; extracting cluster instance characteristics of the cluster instance object; inputting user characteristics of a target user and group instance characteristics of any group instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects; obtaining the matching degree of a target user output by the machine learning model and any group of example objects; and acquiring the group instance object matched with the target user according to the matching degree. According to the method and the device for accurately obtaining the group instance object, the group instance object corresponding to the target user can be accurately obtained, accurate delivery of the group instance object is facilitated, and therefore delivery resources are saved, and user experience is improved.

Description

Group instance object acquisition method and device based on user matching degree

Technical Field

The invention relates to the technical field of data processing, in particular to a group instance object acquisition method and device based on user matching degree.

Background

With the continuous development of science and technology and society, group purchase is increasingly favored by people due to the characteristic of high cost performance. At present, when each platform delivers the group information for users, an indiscriminate delivery mode is generally adopted, namely, the same group information is indiscriminately pushed to all platform users.

However, the inventors found in practice that the following drawbacks exist in the prior art: by adopting the group information throwing mode in the prior art, the throwing group information is not matched with a user, so that the waste of throwing resources is caused, and the user experience is further reduced.

Disclosure of Invention

In view of the foregoing, embodiments of the present invention are provided to provide a method and apparatus for obtaining a clique instance object based on user matching degree, which overcomes the foregoing problems or at least partially solves the foregoing problems.

According to an aspect of the embodiment of the invention, there is provided a clique instance object acquisition method based on user matching degree, including:

extracting user characteristics of a target user from user attribute data of the target user; acquiring group instance data of at least one group instance object with a current state being an operation state, and extracting group instance characteristics of the at least one group instance object from the group instance data; inputting the user characteristics of the target user and the group instance characteristics of any group instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects; obtaining the matching degree of the target user and any group of instance objects output by the machine learning model; and acquiring the group instance object matched with the target user according to the matching degree.

Optionally, the user attribute data includes at least one of the following data: basic attribute data, preference attribute data, consumption capability attribute data, and parameter number attribute data.

Optionally, the bolus instance data includes at least one of: preference level data, group number data, and group rate data.

Optionally, the interaction data of the plurality of historical users with the plurality of historical group instance objects includes at least one of the following data: the method comprises the steps of participating behavior data of a user on a group instance object, grouping result data of the user on the participating group instance object, consumption data of the user on the group instance object which is successfully clustered, and collection behavior data of a commodity corresponding to the non-participating group instance object.

Optionally, before the inputting the user characteristics of the target user and the cluster instance characteristics of any cluster instance object into the pre-trained machine learning model, the method further includes: extracting user characteristics of a plurality of historical users, group instance characteristics of a plurality of historical group instance objects and interaction characteristics of a plurality of historical users and a plurality of historical group instance objects from user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of a plurality of historical users and a plurality of historical group instance objects; for the combination of any historical user and any historical group instance object, generating sample data corresponding to the combination according to the group instance characteristics of the historical group instance object and the interaction characteristics of the historical user and the historical group instance object; and training the generated machine learning model by using the sample data to obtain the trained machine learning model.

Optionally, the training the constructed machine learning model using the sample data to obtain the trained machine learning model further includes: generating positive sample data and negative sample data based on the sample data; wherein the ratio of the positive sample data to the negative sample data satisfies a preset ratio; and training the constructed machine learning model by utilizing the positive sample data and the negative sample data to obtain the trained machine learning model.

Optionally, before the generating the sample data corresponding to the combination according to the user characteristics of the history user, the group instance characteristics of the history group instance object, and the interaction characteristics of the history user and the history group instance object for the combination of any history user and any history group instance object, the method further includes: acquiring a historical user set and a historical group instance object set; wherein any set element in the set of history users corresponds to a history user and any set element in the set of history clique instance objects corresponds to a history clique instance object; carrying out Cartesian product operation on the historical user set and the historical group instance object set, and obtaining a Cartesian set; acquiring a combination of any historical user and any historical group instance object according to the Cartesian set; wherein any set element in the Cartesian set corresponds to a combination of a historic user and a historic clique instance object.

Optionally, after the obtaining the matching degree between the target user and any group of instance objects output by the machine learning model, the method further includes: calculating the clustering success rate of any cluster instance object; correcting the matching degree of the target user and any group of example objects by utilizing the group success rate of any group of example objects so as to obtain a matching degree correction value of the target user and any group of example objects; the obtaining the clique instance object matched with the user according to the matching degree further comprises: and acquiring the group instance object matched with the target user according to the matching degree correction value of the target user and any group instance object.

Optionally, the calculating the clustering success rate of the any cluster instance object further includes: acquiring the number of the agglomeration target persons of any group of example objects and the current number of the reference group; and calculating the clustering success rate of any cluster example object according to the cluster target number of the any cluster example object and the current parameter number.

Optionally, the obtaining the clique instance object matched with the user according to the matching degree further includes: sorting the at least one clique instance object according to the matching degree; and obtaining the clique instance object matched with the target user according to the sorting result.

Optionally, the acquiring the clique instance data of the at least one clique instance object with the current state being the running state further includes: acquiring commodity information of at least one commodity matched with the target user according to the user attribute data of the target user; and acquiring group instance data of at least one group instance object with the current state being an operation state, which corresponds to the commodity information.

Optionally, after the obtaining the clique instance object matched with the target user according to the matching degree, the method further includes: and putting the group instance object in a user terminal of the target user.

According to another aspect of the embodiment of the present invention, there is provided a clique instance object obtaining apparatus based on a user matching degree, including: the user characteristic extraction module is suitable for extracting the user characteristics of the target user from the user attribute data of the target user; the group instance data acquisition module is suitable for acquiring group instance data of at least one group instance object with the current state being the running state; a clique instance feature extraction module adapted to extract clique instance features of the at least one clique instance object from the clique instance data; the input module is suitable for inputting the user characteristics of the target user and the group instance characteristics of any group instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects; the matching degree acquisition module is suitable for acquiring the matching degree of the target user and any group of example objects output by the machine learning model; and the group instance object acquisition module is suitable for acquiring the group instance object matched with the target user according to the matching degree.

Optionally, the apparatus further includes: the history feature extraction module is suitable for extracting user features of a plurality of history users, group instance features of a plurality of history group instance objects and interaction features of a plurality of history users and a plurality of history group instance objects from user attribute data of a plurality of history users, group instance data of a plurality of history group instance objects with running states in a preset history time period and interaction data of a plurality of history users and a plurality of history group instance objects; the sample data generation module is suitable for generating sample data corresponding to any combination of a history user and any history group instance object according to the group instance characteristics of the history group instance object and the interaction characteristics of the history user and the history group instance object; and the training module is suitable for training the generated machine learning model by utilizing the sample data so as to obtain the trained machine learning model.

Optionally, the training module is further adapted to: generating positive sample data and negative sample data based on the sample data; wherein the ratio of the positive sample data to the negative sample data satisfies a preset ratio; and training the constructed machine learning model by utilizing the positive sample data and the negative sample data to obtain the trained machine learning model.

Optionally, the apparatus further includes: the Cartesian operation module is suitable for acquiring a historical user set and a historical group instance object set; wherein any set element in the set of history users corresponds to a history user and any set element in the set of history clique instance objects corresponds to a history clique instance object; carrying out Cartesian product operation on the historical user set and the historical group instance object set, and obtaining a Cartesian set; acquiring a combination of any historical user and any historical group instance object according to the Cartesian set; wherein any set element in the Cartesian set corresponds to a combination of a historic user and a historic clique instance object.

Optionally, the apparatus further includes: the correction module is suitable for calculating the clustering success rate of any cluster instance object; correcting the matching degree of the target user and any group of example objects by utilizing the group success rate of any group of example objects so as to obtain a matching degree correction value of the target user and any group of example objects; the clique instance object acquisition module is further adapted to: and acquiring the group instance object matched with the target user according to the matching degree correction value of the target user and any group instance object.

Optionally, the correction module is further adapted to: acquiring the number of the agglomeration target persons of any group of example objects and the current number of the reference group; and calculating the clustering success rate of any cluster example object according to the cluster target number of the any cluster example object and the current parameter number.

Optionally, the apparatus further includes: the sorting module is suitable for sorting the at least one group instance object according to the matching degree; the clique instance object acquisition module is further adapted to: and obtaining the clique instance object matched with the target user according to the sorting result.

Optionally, the clique instance data acquisition module is further adapted to: acquiring commodity information of at least one commodity matched with the target user according to the user attribute data of the target user; and acquiring group instance data of at least one group instance object with the current state being an operation state, which corresponds to the commodity information.

Optionally, the apparatus further includes: and the delivery module is suitable for delivering the group instance object at the user terminal of the target user.

According to yet another aspect of an embodiment of the present invention, there is provided a computing device including: the device comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete communication with each other through the communication bus; the memory is used for storing at least one executable instruction, and the executable instruction enables the processor to execute the operation corresponding to the group instance object acquisition method based on the user matching degree.

According to still another aspect of the embodiments of the present invention, there is provided a computer storage medium having at least one executable instruction stored therein, the executable instruction causing a processor to perform operations corresponding to the above-described clique instance object obtaining method based on user matching.

According to the method and the device for acquiring the cluster instance object based on the user matching degree, the user characteristics of the target user and the cluster instance characteristics of the cluster instance object are extracted; inputting user characteristics of a target user and group instance characteristics of any group instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects; obtaining the matching degree of a target user output by the machine learning model and any group of example objects; and acquiring the group instance object matched with the target user according to the matching degree. According to the method and the device for accurately obtaining the group instance object, the group instance object corresponding to the target user can be accurately obtained, accurate delivery of the group instance object is facilitated, and therefore delivery resources are saved, and user experience is improved.

The foregoing description is only an overview of the present invention, and is intended to be implemented in accordance with the teachings of the present invention in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present invention more readily apparent.

Drawings

Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to designate like parts throughout the figures. In the drawings:

fig. 1 is a flow chart illustrating a method for obtaining a clique instance object based on user matching according to an embodiment of the present invention;

fig. 2 is a flow chart of a method for obtaining a clique instance object based on user matching degree according to a second embodiment of the present invention;

fig. 3 is a flow chart of a method for obtaining a clique instance object based on user matching degree according to a third embodiment of the present invention;

fig. 4 is a schematic structural diagram of a clique instance object obtaining device based on user matching degree according to a fourth embodiment of the present invention;

Fig. 5 shows a schematic structural diagram of a computing device according to a sixth embodiment of the present invention.

Detailed Description

Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

Example 1

Fig. 1 is a flow chart illustrating a method for obtaining a clique instance object based on user matching according to an embodiment of the present invention. The group instance object obtaining method provided by the embodiment can be applied to service platforms of various industries, for example, a local life service platform, an online electronic commerce platform, a take-away platform and the like. Specifically, the group instance object obtaining method provided in the present embodiment can be executed by a computing device having a corresponding computing capability, and the specific type of the computing device and the like are not limited in the present embodiment.

As shown in fig. 1, the method includes:

Step S110: and extracting the user characteristics of the target user from the user attribute data of the target user.

In order to be able to put matched clique instance objects for users, the present embodiment obtains user attribute data of the targeted users. Wherein the user attribute data includes at least one of the following data: basic attribute data, preference attribute data, consumption capability attribute data, and parameter number attribute data. Specifically, the basic attribute data may include data of age, occupation, and/or sex, etc., which may be obtained through basic information or the like input by the user when registering the platform; the preference attribute data is specifically data reflecting the preference habit of the user, the data can be obtained by historical consumption and browsing data of the user, for example, the commodity category preference of the user is determined according to the commodity category information with the greatest consumption of the historical user; the consumption capability attribute data may include consumption capability data of a user history and/or current consumption capability data of the user, wherein the consumption capability data of the user history may be obtained according to data such as consumption records of the user history, and the current consumption capability data of the user may be obtained according to account information of the user in a preset platform; in addition, the parameter group times attribute data of the user can be obtained according to the historical times information of the user participating in the group spelling activities. In short, the specific acquisition mode and the like of the user attribute data are not limited in this embodiment.

Further, after the user attribute data of the target user is obtained, the user characteristics of the target user are extracted from the user attribute data, and the user characteristics can intensively reflect the relevant characteristics of the target user. The specific feature extraction method is not limited in this embodiment. For example, if the obtained user attribute data includes structured data, the target user feature may be directly extracted according to the related attribute information of the structured data table, for example, the gender data of the target user may be directly extracted from the data table including the user ID and the gender attribute through the related query statement; if the obtained user attribute data contains unstructured data (such as long text data, etc.), word segmentation processing can be performed on the unstructured data first, so as to obtain a plurality of text word segments; and further extracting feature words from the text segmentation words, and taking the feature words as user features of the target user.

Step S120: and acquiring the cluster instance data of at least one cluster instance object with the current state being the running state, and extracting the cluster instance characteristics of the at least one cluster instance object from the cluster instance data.

The cluster instance object obtained in this step is specifically a cluster instance object currently in an operation state. In the implementation process, the effective operation time period and the current time of at least one group instance object can be obtained, whether the effective operation time period of any group instance object contains the current time or not is judged, and if yes, group instance data of the group instance object are obtained.

Specifically, the clique instance data includes at least one of the following data: preference level data, group number data, and group rate data. Wherein the preference degree data further comprises preference proportion and/or preference amount of the group instance object; the group number data comprises a target group number corresponding to the group instance object (wherein, when the group number reaches the target group number, the group is indicated) and the current group number; the cluster rate data includes a cluster ratio corresponding to the current cluster instance object.

Further, a clique instance feature of at least one clique instance object is extracted from the obtained clique instance data, the clique instance feature being capable of reflecting specific characteristics of the clique instance object. The specific feature extraction manner in this embodiment is not limited, for example, reference may be made to the description of the corresponding portion in step S110, and the detailed description of this step is omitted here.

Step S130, inputting user characteristics of a target user and group instance characteristics of any group instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects.

In this embodiment, an organic machine learning model is built in advance, and the built machine learning model is trained by using the obtained history data, so as to obtain a trained machine learning model. Wherein, the history data specifically includes: user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects. Optionally, the interaction data of the plurality of historical users with the plurality of historical group instance objects includes at least one of: the method comprises the steps of participating in behavior data of a group instance object by a user (the data comprises a participating group instance object or an un-participating group instance object), grouping result data of the participating group instance object by the user (the data comprises successful grouping of the participating group instance object or failed grouping of the participating group instance object), consumption data of the group instance object by the user (the data comprises successful consumption of the group instance object which is successful in grouping, unsuccessful consumption of the group instance object which is successful in grouping), and collection behavior data of commodities corresponding to the un-participating group instance object by the user (the commodities corresponding to the un-participating group instance object have collection behaviors or the commodities corresponding to the un-participating group instance object do not have collection behaviors). The user's participation behavior data of the group instance object may be that the history user has participated in the group instance object or has not participated in the group instance object; the group result data of the participating group instance object may be consumed or not consumed by the user.

According to the embodiment, the machine learning model is trained by utilizing the user attribute data of the historical user, the group instance data of the historical group instance object and the interaction data of a plurality of historical users and a plurality of historical group instance objects, so that the machine learning model can acquire parameter group achievements corresponding to different user characteristics and group instance characteristics, and the preference degree of different users on different group instances can be accurately acquired through the parameter group achievements, so that the matching degree of a target user and any group instance object can be accurately predicted.

In an alternative embodiment, when the number of clique instance objects whose current state is the running state is large, data input to the machine learning model is avoided from being missed. In this embodiment, a current group instance object set may be generated, and a cartesian product operation is performed on a user set including a target user and the current group instance object set, so as to obtain a cartesian set, and a combination of the target user and any group instance object is determined according to the generated cartesian set, where each set element in the cartesian set corresponds to a combination of the target user and one group instance object. By adopting the mode, the combination of the target user and the group instance object with the current state being the running state can be accurately obtained, so that missing of inputting data to the machine learning model is avoided; moreover, all combinations of the target user and the group instance object with the current state being the running state can be rapidly determined through single operation, so that the acquisition efficiency of the combination formed by the target user and different group instance objects can be effectively improved, and the improvement of the overall implementation efficiency of the method is facilitated.

Step S140, the matching degree of the target user output by the machine learning model and any group of instance objects is obtained.

The matching degree of the target user and any group of instance objects can be accurately obtained through the machine learning model. The higher the matching degree between the target user and the group instance object, the higher the preference degree of the target user to the group instance object, and the easier the target user participates in or consumes the group instance object.

And step S150, acquiring a group instance object matched with the target user according to the matching degree.

After the degree of matching of the target user with any of the clique instance objects is obtained, the clique instance object that matches the target user may be determined further based on the obtained degree of matching. For example, at least one clique instance object may be ranked according to the degree of matching, and clique instance objects that match the target user may be obtained according to the ranking result (e.g., the clique instance object n-th in the ranking rank may be used as the clique instance object that matches the target user); or, taking the group instance object with the matching degree larger than the preset matching degree as the group instance object matched with the target user.

Optionally, after the group instance object matched with the target user is obtained according to the matching degree, the group instance object may be further launched at the user terminal of the target user, and the specific launching form is not limited in this embodiment.

Therefore, in this embodiment, the machine learning model is trained by using the historical data, specifically, the user attribute data of the historical user, the group instance data of the historical group instance object, and the interaction data of the plurality of historical users and the plurality of historical group instance objects, so that the trained machine learning model can accurately obtain the matching degree of the target user and the group instance object; and further, the group instance object matched with the target user is determined according to the matching degree, thereby being convenient for realizing the accurate delivery of the group instance object, avoiding the defects of user experience reduction, system transmission and other delivery resource waste caused by delivering the group instance object which is not interested by the user in the prior art, and further being beneficial to the promotion of the user experience and saving the delivery resource.

Example two

Fig. 2 is a flow chart of a clique instance object obtaining method based on user matching degree according to a second embodiment of the present invention. The method for obtaining the clique instance object provided in the present embodiment is further optimized for the method for obtaining the clique instance object in the first embodiment.

As shown in fig. 2, the method includes:

step S210: a machine learning model is generated.

The specific type of the machine learning model and the like are not limited in this embodiment. For example, a machine learning model of multiple neural network layers may be employed, where the machine learning model includes an input layer, at least one fully connected layer, and an output layer. The input layer user receives input data, the full-connection layer user processes the received input data, and the output layer is used for outputting a result.

Step S220: and generating sample data according to the user attribute data of the plurality of historical users, group instance data of the plurality of historical group instance objects with the running states in the preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects.

In this embodiment, the historical data is specifically utilized to generate corresponding sample data according to user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with states being running states in a preset historical period, and interaction data of the plurality of historical users and the plurality of historical group instance objects.

Specifically, first, user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with states of running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects are obtained. In this embodiment, the method for acquiring the history data is not limited.

Further, user characteristics of a plurality of history users, group instance characteristics of a plurality of history group instance objects and interaction characteristics of a plurality of history users and a plurality of history group instance objects are extracted from the obtained history data. The specific extraction manner in this embodiment may refer to the description of the corresponding portion in the first embodiment, which is not described herein. In an alternative implementation manner, in order to facilitate subsequent training of the machine learning model, the interaction features extracted in this embodiment are specifically different interaction degree identifiers, for example, identifier 2 corresponds to that the user successfully consumes the cluster instance object that is clustered successfully, which indicates that the preference degree of the user on the cluster instance object is the highest; the mark 1 corresponds to that the user does not successfully consume the cluster instance object with successful cluster, or the cluster instance object participated by the user does not successfully cluster, or the user collects the commodity corresponding to the cluster instance object, which indicates that the preference degree of the user to the cluster instance object is higher; the identification 0 corresponds to the group instance object which the user does not participate in, and the commodity corresponding to the group instance object which is not collected indicates that the preference degree of the user to the group instance object is the weakest.

Still further, for a combination of any historical user and any historical clique instance object, sample data corresponding to the combination is generated according to user characteristics of the historical user, clique instance characteristics of the historical clique instance object, and interaction characteristics of the historical user and the historical clique instance object. That is, in this embodiment, a combination of a history user and a history group instance object corresponds to a piece of sample data, where the sample data includes a user feature of the history user, a group instance feature of the history group instance object, and an interaction feature of the history user and the history group instance object.

In an alternative embodiment, to improve the prediction accuracy of the machine learning model, the machine learning model may be trained using a large amount of sample data. When model training is performed using a large amount of sample data, a large number of combinations of historical users and historical clique instance objects need to be determined. In this embodiment, a history user set and a history group instance object set may be obtained; wherein any set element in the set of history users corresponds to a history user and any set element in the set of history group instance objects corresponds to a history group instance object; carrying out Cartesian product operation on the historical user set and the historical group instance object set, and obtaining a Cartesian set; acquiring a combination of a history user and a history group instance object according to a Cartesian set; wherein any set element in the Cartesian set corresponds to a combination of a historic user and a historic clique instance object. By adopting the implementation mode, the combination of the history user and the history group instance object can be quickly and accurately obtained from a large amount of data, so that the execution efficiency of the method is improved.

In an alternative embodiment, after generating sample data corresponding to each combination of historical user and historical clique instance objects, sample balancing processing may be further performed on the obtained sample data. Specifically, in the actual implementation process, because there is a large difference in the number of users thrown by different historical group instance objects (for example, the number of users thrown by the historical group instance object a is 100000, and the number of users thrown by the historical group instance object B is 100), the generated sample data has a large difference in the sample data corresponding to different historical group instance objects, so that the subsequent machine learning model easily has a large difference in the prediction precision of different historical group instance objects, and especially, the defect that the prediction precision of the historical group instance objects of small-capacity sample data is low is easily caused. In order to avoid this disadvantage, the present embodiment performs sample balancing processing on the obtained sample data. Namely, aiming at a historical group instance object of large sample data, sampling is carried out by adopting a lower sampling frequency; and for the historical group instance object of the small sample data, the higher sampling frequency is adopted for sampling, so that the sample data corresponding to different historical group instance objects in the sample data after the balance processing have the same or less difference.

In the embodiment, the execution sequence of the step S210 and the step S220 is not limited, and the two may be executed in parallel or sequentially.

Step S230: and training the generated machine learning model by using the sample data to obtain a trained machine learning model.

Specifically, positive sample data as well as negative sample data may be generated based on the sample data; wherein the ratio of the positive sample data to the negative sample data satisfies a preset ratio (e.g., the ratio of the positive sample data to the negative sample data is 3:1); and training the constructed machine learning model by utilizing the positive sample data and the negative sample data to obtain a trained machine learning model.

The specific machine learning model training method is not limited in this embodiment. For example, XGBoost (eXtreme Gradient Boosting, extreme gradient lifting) algorithm, or random forest algorithm may be employed for training of machine learning models, and so on.

Step S240: and extracting the user characteristics of the target user from the user attribute data of the target user.

Step S250: and acquiring the cluster instance data of at least one cluster instance object with the current state being the running state, and extracting the cluster instance characteristics of the at least one cluster instance object from the cluster instance data.

Step S260: and inputting the user characteristics of the target user and the cluster instance characteristics of any cluster instance object into a pre-trained machine learning model.

Step S270: and acquiring the matching degree of the target user and any group instance object output by the machine learning model, and acquiring the group instance object matched with the target user according to the matching degree.

The specific implementation process of steps S240 to S270 may refer to the description of the corresponding parts in the first embodiment, which is not described herein.

Therefore, the embodiment trains the machine learning model by utilizing the user attribute data of the historical user, the group instance data of the historical group instance object and the interaction data of a plurality of historical users and a plurality of historical group instance objects, so that the machine learning model can acquire parameter group achievements corresponding to the combination of different user characteristics and group instance characteristics, and can accurately acquire the preference degree of different users to different group instances through the parameter group achievements, thereby accurately predicting the matching degree of a target user and any group instance object; in addition, in the sample data processing process, the combination of each history user and each history group instance object can be rapidly and accurately determined by adopting a Cartesian product operation mode, so that the method is beneficial to improving the execution efficiency; moreover, the embodiment further ensures the consistency of the prediction precision of the machine learning model on different clique example objects through the balance processing of the sample data.

Example III

Fig. 3 is a flow chart illustrating a method for obtaining a clique instance object based on user matching degree according to a third embodiment of the present invention. The method for obtaining the clique instance object provided in the present embodiment is further optimized for the method for obtaining the clique instance object in the first embodiment and/or the second embodiment.

As shown in fig. 3, the method includes:

step S310: and extracting the user characteristics of the target user from the user attribute data of the target user.

Step S320: and acquiring the cluster instance data of at least one cluster instance object with the current state being the running state, and extracting the cluster instance characteristics of the at least one cluster instance object from the cluster instance data.

In an alternative embodiment, commodity information of at least one commodity matched with the target user can be obtained according to the user attribute data of the target user; and further acquiring group instance data of at least one group instance object with the current state being an operation state corresponding to the commodity information. Therefore, the embodiment only acquires the group instance data of the group instance object of the commodity interested by the user, thereby reducing the data processing amount of the subsequent steps and further avoiding the waste of system resources; and the finally obtained clique instance object matched with the target user is made to correspond to the actual preference degree of the target user.

Step S330: and inputting the user characteristics of the target user and the cluster instance characteristics of any cluster instance object into a pre-trained machine learning model.

The pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects.

Step S340: and obtaining the matching degree of the target user output by the machine learning model and any group of instance objects.

The specific implementation process of step S310 to step S320 may refer to the description of the corresponding parts in the first embodiment, which is not described herein.

Step S350: calculating the clustering success rate of any cluster of instance objects, and correcting the matching degree of the target user and any cluster of instance objects by using the clustering success rate of any cluster of instance objects to obtain a matching degree correction value of the target user and any cluster of instance objects.

In this embodiment, after obtaining the matching degree between the target user and any group instance object output by the machine learning model, in order to further improve the finally determined actual matching degree between the group instance object and the target user, the matching degree between the target user and any group instance object is corrected through this step.

In a specific correction process, the success rate of clustering of any cluster instance object is calculated first. The clustering success rate of any cluster instance object is calculated specifically as follows: acquiring the number of the agglomeration target persons of any group of example objects and the current number of the reference group; the clustering success rate of any one cluster example object is calculated according to the clustering target number of any cluster example object and the current parameter number, for example, the clustering success rate of the cluster example object can be calculated according to the ratio of the difference value of the clustering target number and the current parameter number to the clustering target number, and the higher the ratio is, the lower the clustering success rate is.

Further, the matching degree of the target user and any group of example objects is corrected by utilizing the group success rate of any group of example objects, so as to obtain a matching degree correction value of the target user and any group of example objects. Wherein the matching degree correction value of the target user and any group instance object is positively related to the grouping success rate of the group instance object. For example, the matching correction value for the target user to any clique instance object may be calculated using the following formula:

wherein P' is a matching degree correction value; p is the matching degree before correction; alpha is the ratio of the difference value of the number of the agglomeration targets and the current number of the reference agglomeration targets to the number of the agglomeration targets.

Step S360: and acquiring the group instance object matched with the target user according to the matching degree correction value of the target user and any group instance object.

Therefore, after the matching degree of the target user and any group of example objects output by the machine learning model is obtained, the matching degree of the target user and any group of example objects is further corrected, so that the corrected matching degree can more truly reflect the preference degree of the target user on the group of example objects, and the improvement of the clustering rate is facilitated; in addition, the group instance object corresponding to the target user obtained in the embodiment is the group instance object of the commodity of interest to the user, so that the calculated amount of the method is reduced, meanwhile, the system resource can be effectively saved, and the finally obtained group instance object is matched with the actual preference degree of the target user.

Example IV

Fig. 4 is a schematic structural diagram of a clique instance object obtaining device based on user matching degree according to a fourth embodiment of the present invention. As shown in fig. 4, the apparatus includes: a user feature extraction module 41, a cluster instance data acquisition module 42, a cluster instance feature extraction module 43, an input module 44, a matching degree acquisition module 45 and a cluster instance object acquisition module 46.

A user feature extraction module 41 adapted to extract user features of a target user from user attribute data of the target user; a clique instance data acquisition module 42 adapted to acquire clique instance data of at least one clique instance object whose current state is an operational state; a clique instance feature extraction module 43 adapted to extract clique instance features of the at least one clique instance object from the clique instance data; an input module 44 adapted to input user features of the target user and clique instance features of any clique instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects; the matching degree obtaining module 45 is adapted to obtain the matching degree between the target user and any group of example objects output by the machine learning model; a clique instance object acquisition module 46 is adapted to acquire clique instance objects matching the target user based on the degree of matching.

Optionally, the apparatus further includes: the Cartesian operation module is suitable for acquiring a historical user set and a historical group instance object set; wherein any set element in the set of history users corresponds to a history user and any set element in the set of history clique instance objects corresponds to a history clique instance object; carrying out Cartesian product operation on the historical user set and the historical group instance object set, and obtaining a Cartesian set; acquiring a combination of a history user and a history group instance object according to the Cartesian set; wherein any set element in the Cartesian set corresponds to a combination of a historic user and a historic clique instance object.

Optionally, the apparatus further includes: the correction module is suitable for calculating the clustering success rate of any cluster instance object; correcting the matching degree of the target user and any group of example objects by utilizing the group success rate of any group of example objects so as to obtain a matching degree correction value of the target user and any group of example objects;

The clique instance object acquisition module matching is further adapted to: and acquiring the group instance object matched with the target user according to the matching degree correction value of the target user and any group instance object.

Optionally, the correction module is further adapted to: acquiring the number of the agglomeration target persons of any group of example objects and the current number of the reference group;

and calculating the clustering success rate of any cluster example object according to the cluster target number of the any cluster example object and the current parameter number.

Optionally, the apparatus further includes:

the sorting module is suitable for sorting the at least one group instance object according to the matching degree;

the clique instance object acquisition module is further adapted to: and obtaining the clique instance object matched with the target user according to the sorting result.

The specific implementation process of each module in this embodiment may refer to the description in the corresponding method embodiment, and this embodiment is not described herein.

Therefore, the method and the device can accurately acquire the group instance object corresponding to the target user, and are convenient to realize accurate delivery of the group instance object, so that delivery resources are saved, and user experience is improved.

Example five

Embodiments of the present invention provide a non-transitory computer storage medium storing at least one executable instruction for performing the method of any of the method embodiments described above.

The executable instructions may be particularly useful for causing a processor to: extracting user characteristics of a target user from user attribute data of the target user; acquiring group instance data of at least one group instance object with a current state being an operation state, and extracting group instance characteristics of the at least one group instance object from the group instance data; inputting the user characteristics of the target user and the group instance characteristics of any group instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects; obtaining the matching degree of the target user and any group of instance objects output by the machine learning model; and acquiring the group instance object matched with the target user according to the matching degree.

In an alternative embodiment, the user attribute data includes at least one of the following: basic attribute data, preference attribute data, consumption capability attribute data, and parameter number attribute data.

In an alternative embodiment, the bolus instance data includes at least one of the following data: preference level data, group number data, and group rate data.

In an alternative embodiment, the interaction data of the plurality of historical users with the plurality of historical group instance objects includes at least one of the following data: the method comprises the steps of participating behavior data of a user on a group instance object, grouping result data of the user on the participating group instance object, consumption data of the user on the group instance object which is successfully clustered, and collection behavior data of a commodity corresponding to the non-participating group instance object.

In an alternative embodiment, the executable instructions may be specifically configured to cause a processor to: extracting user characteristics of a plurality of historical users, group instance characteristics of a plurality of historical group instance objects and interaction characteristics of a plurality of historical users and a plurality of historical group instance objects from user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of a plurality of historical users and a plurality of historical group instance objects; for the combination of any historical user and any historical group instance object, generating sample data corresponding to the combination according to the group instance characteristics of the historical group instance object and the interaction characteristics of the historical user and the historical group instance object; and training the generated machine learning model by using the sample data to obtain the trained machine learning model.

In an alternative embodiment, the executable instructions may be specifically configured to cause a processor to: generating positive sample data and negative sample data based on the sample data; wherein the ratio of the positive sample data to the negative sample data satisfies a preset ratio; and training the constructed machine learning model by utilizing the positive sample data and the negative sample data to obtain the trained machine learning model.

In an alternative embodiment, the executable instructions may be specifically configured to cause a processor to: acquiring a historical user set and a historical group instance object set; wherein any set element in the set of history users corresponds to a history user and any set element in the set of history clique instance objects corresponds to a history clique instance object; carrying out Cartesian product operation on the historical user set and the historical group instance object set, and obtaining a Cartesian set; acquiring a combination of any historical user and any historical group instance object according to the Cartesian set; wherein any set element in the Cartesian set corresponds to a combination of a historic user and a historic clique instance object.

In an alternative embodiment, the executable instructions may be specifically configured to cause a processor to: calculating the clustering success rate of any cluster instance object; correcting the matching degree of the target user and any group of example objects by utilizing the group success rate of any group of example objects so as to obtain a matching degree correction value of the target user and any group of example objects; and acquiring the group instance object matched with the target user according to the matching degree correction value of the target user and any group instance object.

In an alternative embodiment, the executable instructions may be specifically configured to cause a processor to: acquiring the number of the agglomeration target persons of any group of example objects and the current number of the reference group; and calculating the clustering success rate of any cluster example object according to the cluster target number of the any cluster example object and the current parameter number.

In an alternative embodiment, the executable instructions may be specifically configured to cause a processor to: sorting the at least one clique instance object according to the matching degree; and obtaining the clique instance object matched with the target user according to the sorting result.

In an alternative embodiment, the executable instructions may be specifically configured to cause a processor to: acquiring commodity information of at least one commodity matched with the target user according to the user attribute data of the target user; and acquiring group instance data of at least one group instance object with the current state being an operation state, which corresponds to the commodity information.

In an alternative embodiment, the executable instructions may be specifically configured to cause a processor to: and putting the group instance object in a user terminal of the target user.

Example six

Fig. 5 shows a schematic structural diagram of a computing device according to a sixth embodiment of the present invention, which is not limited to the specific implementation of the computing device according to the embodiments of the present invention.

As shown in fig. 5, the computing device may include: a processor 502, a communication interface (Communications Interface) 504, a memory 506, and a communication bus 508. Wherein: processor 502, communication interface 504, and memory 506 communicate with each other via communication bus 508. A communication interface 504 for communicating with network elements of other devices, such as clients or other servers. The processor 502 is configured to execute the program 510, and may specifically perform relevant steps in the method embodiments described above. In particular, program 510 may include program code including computer-operating instructions. The processor 502 may be a central processing unit CPU, or a specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors included by the computing device may be the same type of processor, such as one or more CPUs; but may also be different types of processors such as one or more CPUs and one or more ASICs.

A memory 506 for storing a program 510. Memory 506 may comprise high-speed RAM memory or may also include non-volatile memory (non-volatile memory), such as at least one disk memory.

The program 510 may be specifically operable to cause the processor 502 to: extracting user characteristics of a target user from user attribute data of the target user; acquiring group instance data of at least one group instance object with a current state being an operation state, and extracting group instance characteristics of the at least one group instance object from the group instance data; inputting the user characteristics of the target user and the group instance characteristics of any group instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects; obtaining the matching degree of the target user and any group of instance objects output by the machine learning model; and acquiring the group instance object matched with the target user according to the matching degree.

In an alternative embodiment, program 510 may be specifically configured to cause processor 502 to: extracting user characteristics of a plurality of historical users, group instance characteristics of a plurality of historical group instance objects and interaction characteristics of a plurality of historical users and a plurality of historical group instance objects from user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of a plurality of historical users and a plurality of historical group instance objects; for the combination of any historical user and any historical group instance object, generating sample data corresponding to the combination according to the group instance characteristics of the historical group instance object and the interaction characteristics of the historical user and the historical group instance object; and training the generated machine learning model by using the sample data to obtain the trained machine learning model.

In an alternative embodiment, program 510 may be specifically configured to cause processor 502 to: generating positive sample data and negative sample data based on the sample data; wherein the ratio of the positive sample data to the negative sample data satisfies a preset ratio; and training the constructed machine learning model by utilizing the positive sample data and the negative sample data to obtain the trained machine learning model.

In an alternative embodiment, program 510 may be specifically configured to cause processor 502 to: acquiring a historical user set and a historical group instance object set; wherein any set element in the set of history users corresponds to a history user and any set element in the set of history clique instance objects corresponds to a history clique instance object; carrying out Cartesian product operation on the historical user set and the historical group instance object set, and obtaining a Cartesian set; acquiring a combination of any historical user and any historical group instance object according to the Cartesian set; wherein any set element in the Cartesian set corresponds to a combination of a historic user and a historic clique instance object.

In an alternative embodiment, program 510 may be specifically configured to cause processor 502 to: calculating the clustering success rate of any cluster instance object; correcting the matching degree of the target user and any group of example objects by utilizing the group success rate of any group of example objects so as to obtain a matching degree correction value of the target user and any group of example objects; and acquiring the group instance object matched with the target user according to the matching degree correction value of the target user and any group instance object.

In an alternative embodiment, program 510 may be specifically configured to cause processor 502 to: acquiring the number of the agglomeration target persons of any group of example objects and the current number of the reference group; and calculating the clustering success rate of any cluster example object according to the cluster target number of the any cluster example object and the current parameter number.

In an alternative embodiment, program 510 may be specifically configured to cause processor 502 to: sorting the at least one clique instance object according to the matching degree; and obtaining the clique instance object matched with the target user according to the sorting result.

In an alternative embodiment, program 510 may be specifically configured to cause processor 502 to: acquiring commodity information of at least one commodity matched with the target user according to the user attribute data of the target user; and acquiring group instance data of at least one group instance object with the current state being an operation state, which corresponds to the commodity information.

In an alternative embodiment, program 510 may be specifically configured to cause processor 502 to: and putting the group instance object in a user terminal of the target user.

The algorithms or displays presented herein are not inherently related to any particular computer, virtual system, or other apparatus. Various general-purpose systems may also be used with the teachings herein. The required structure for a construction of such a system is apparent from the description above. In addition, embodiments of the present invention are not directed to any particular programming language. It will be appreciated that the teachings of the present invention described herein may be implemented in a variety of programming languages, and the above description of specific languages is provided for disclosure of enablement and best mode of the present invention.

In the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.

Similarly, it should be appreciated that in the above description of exemplary embodiments of the invention, various features of the embodiments of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be construed as reflecting the intention that: i.e., the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.

Those skilled in the art will appreciate that the modules in the apparatus of the embodiments may be adaptively changed and disposed in one or more apparatuses different from the embodiments. The modules or units or components of the embodiments may be combined into one module or unit or component and, furthermore, they may be divided into a plurality of sub-modules or sub-units or sub-components. Any combination of all features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or units of any method or apparatus so disclosed, may be used in combination, except insofar as at least some of such features and/or processes or units are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings), may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.

Furthermore, those skilled in the art will appreciate that while some embodiments herein include some features but not others included in other embodiments, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the following claims, any of the claimed embodiments can be used in any combination.

Various component embodiments of the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art will appreciate that some or all of the functionality of some or all of the components according to embodiments of the present invention may be implemented in practice using a microprocessor or Digital Signal Processor (DSP). The present invention can also be implemented as an apparatus or device program (e.g., a computer program and a computer program product) for performing a portion or all of the methods described herein. Such a program embodying the present invention may be stored on a computer readable medium, or may have the form of one or more signals. Such signals may be downloaded from an internet website, provided on a carrier signal, or provided in any other form.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second, third, etc. do not denote any order. These words may be interpreted as names. The steps in the above embodiments should not be construed as limiting the order of execution unless specifically stated.

Claims

1. A group instance object acquisition method based on user matching degree comprises the following steps:

extracting user characteristics of a target user from user attribute data of the target user;

acquiring group instance data of at least one group instance object with a current state being an operation state, and extracting group instance characteristics of the at least one group instance object from the group instance data;

inputting the user characteristics of the target user and the group instance characteristics of any group instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects;

obtaining the matching degree of the target user and any group of instance objects output by the machine learning model;

and acquiring the group instance object matched with the target user according to the matching degree.

2. The method of claim 1, wherein the user attribute data comprises at least one of:

Basic attribute data, preference attribute data, consumption capability attribute data, and parameter number attribute data.

3. The method of claim 1, wherein the clique instance data comprises at least one of:

preference level data, group number data, and group rate data.

4. The method of claim 1, wherein the interaction data of the plurality of historical users with the plurality of historical clique instance objects comprises at least one of:

the method comprises the steps of participating behavior data of a user on a group instance object, grouping result data of the user on the participating group instance object, consumption data of the user on the group instance object which is successfully clustered, and collection behavior data of a commodity corresponding to the non-participating group instance object.

5. The method of any of claims 1-4, wherein prior to the inputting the user characteristics of the target user and the clique instance characteristics of any clique instance object into a pre-trained machine learning model, the method further comprises:

extracting user characteristics of a plurality of historical users, group instance characteristics of a plurality of historical group instance objects and interaction characteristics of a plurality of historical users and a plurality of historical group instance objects from user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of a plurality of historical users and a plurality of historical group instance objects;

For the combination of any historical user and any historical group instance object, generating sample data corresponding to the combination according to the group instance characteristics of the historical group instance object and the interaction characteristics of the historical user and the historical group instance object;

and training the generated machine learning model by using the sample data to obtain the trained machine learning model.

6. The method of claim 5, wherein the training the constructed machine learning model with the sample data to obtain the trained machine learning model further comprises:

generating positive sample data and negative sample data based on the sample data; wherein the ratio of the positive sample data to the negative sample data satisfies a preset ratio;

and training the constructed machine learning model by utilizing the positive sample data and the negative sample data to obtain the trained machine learning model.

7. The method of claim 5, wherein, prior to generating the sample data corresponding to any combination of a history user and any history clique instance object based on user characteristics of the history user, clique instance characteristics of the history clique instance object, and interaction characteristics of the history user and the history clique instance object, the method further comprises:

Acquiring a historical user set and a historical group instance object set; wherein any set element in the set of history users corresponds to a history user and any set element in the set of history clique instance objects corresponds to a history clique instance object;

carrying out Cartesian product operation on the historical user set and the historical group instance object set, and obtaining a Cartesian set;

acquiring a combination of any historical user and any historical group instance object according to the Cartesian set; wherein any set element in the Cartesian set corresponds to a combination of a historic user and a historic clique instance object.

8. The method of any of claims 1-4, wherein after the obtaining the degree of matching of the target user of the machine learning model output to any clique instance object, the method further comprises:

calculating the clustering success rate of any cluster instance object;

correcting the matching degree of the target user and any group of example objects by utilizing the group success rate of any group of example objects so as to obtain a matching degree correction value of the target user and any group of example objects;

The obtaining the clique instance object matched with the user according to the matching degree further comprises:

and acquiring the group instance object matched with the target user according to the matching degree correction value of the target user and any group instance object.

9. The method of claim 8, wherein the calculating the cluster success rate for the any cluster instance object further comprises:

acquiring the number of the agglomeration target persons of any group of example objects and the current number of the reference group;

10. The method of any of claims 1-4, wherein the obtaining a clique instance object matching the user according to the degree of matching further comprises:

sorting the at least one clique instance object according to the matching degree;

and obtaining the clique instance object matched with the target user according to the sorting result.

11. The method of any of claims 1-4, wherein the obtaining clique instance data of at least one clique instance object whose current state is a running state further comprises:

Acquiring commodity information of at least one commodity matched with the target user according to the user attribute data of the target user;

and acquiring group instance data of at least one group instance object with the current state being an operation state, which corresponds to the commodity information.

12. The method of any of claims 1-4, wherein after the obtaining a clique instance object matching the target user according to the degree of matching, the method further comprises:

and putting the group instance object in a user terminal of the target user.

13. A clique instance object acquisition device based on user matching, comprising:

the user characteristic extraction module is suitable for extracting the user characteristics of the target user from the user attribute data of the target user;

the group instance data acquisition module is suitable for acquiring group instance data of at least one group instance object with the current state being the running state;

a clique instance feature extraction module adapted to extract clique instance features of the at least one clique instance object from the clique instance data;

the input module is suitable for inputting the user characteristics of the target user and the group instance characteristics of any group instance object into a pre-trained machine learning model; the pre-trained machine learning model is obtained by training based on user attribute data of a plurality of historical users, group instance data of a plurality of historical group instance objects with running states in a preset historical time period and interaction data of the plurality of historical users and the plurality of historical group instance objects;

The matching degree acquisition module is suitable for acquiring the matching degree of the target user and any group of example objects output by the machine learning model;

and the group instance object acquisition module is suitable for acquiring the group instance object matched with the target user according to the matching degree.

14. The apparatus of claim 13, wherein the user attribute data comprises at least one of:

15. The apparatus of claim 13, wherein the clique instance data comprises at least one of:

preference level data, group number data, and group rate data.

16. The apparatus of claim 13, wherein the interaction data of the plurality of historical users with the plurality of historical clique instance objects comprises at least one of:

17. The apparatus of any of claims 13-16, wherein the apparatus further comprises:

the history feature extraction module is suitable for extracting user features of a plurality of history users, group instance features of a plurality of history group instance objects and interaction features of a plurality of history users and a plurality of history group instance objects from user attribute data of a plurality of history users, group instance data of a plurality of history group instance objects with running states in a preset history time period and interaction data of a plurality of history users and a plurality of history group instance objects;

the sample data generation module is suitable for generating sample data corresponding to any combination of a history user and any history group instance object according to the group instance characteristics of the history group instance object and the interaction characteristics of the history user and the history group instance object;

and the training module is suitable for training the generated machine learning model by utilizing the sample data so as to obtain the trained machine learning model.

18. The apparatus of claim 17, wherein the training module is further adapted to:

19. The apparatus of claim 17, wherein the apparatus further comprises:

the Cartesian operation module is suitable for acquiring a historical user set and a historical group instance object set; wherein any set element in the set of history users corresponds to a history user and any set element in the set of history clique instance objects corresponds to a history clique instance object;

20. The apparatus of any of claims 13-16, wherein the apparatus further comprises:

the correction module is suitable for calculating the clustering success rate of any cluster instance object;

The clique instance object acquisition module is further adapted to: and acquiring the group instance object matched with the target user according to the matching degree correction value of the target user and any group instance object.

21. The apparatus of claim 20, wherein the correction module is further adapted to: acquiring the number of the agglomeration target persons of any group of example objects and the current number of the reference group;

22. The apparatus of any of claims 13-16, wherein the apparatus further comprises:

23. The apparatus of any of claims 13-16, wherein the bolus instance data obtaining module is further adapted to:

24. The apparatus of any of claims 13-16, wherein the apparatus further comprises: and the delivery module is suitable for delivering the group instance object at the user terminal of the target user.

25. A computing device, comprising: the device comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete communication with each other through the communication bus;

the memory is configured to store at least one executable instruction, where the executable instruction causes the processor to perform operations corresponding to the method for obtaining a clique instance object based on user matching as claimed in any one of claims 1 to 12.

26. A computer storage medium having stored therein at least one executable instruction for causing a processor to perform operations corresponding to the user matching based clique instance object retrieval method of any of claims 1-12.